The primary goal of this project is to develop and evaluate new statistical methodology for cell- type-specific analysis of genomic data, discerning intratumor heterogeneity (ITH), and joint modeling of multivariate longitudinal and time-to-event data. The proposed three aims, each having two sub-aims, have major clinical relevance and have potential for major clinical impact. The proposed three aims present a unified approach within the joint modeling framework for mixed types of genomic data, longitudinal biomarkers, and time-to-event data for a better understanding of epigenome-wide associations, intratumor heterogeneity, and associations between genomic and/or longitudinal biomarkers and time-to-event outcomes useful in cancer research. All findings are preliminary and no manuscripts have been published on any of the proposed aims thus far. The overall unifying theme of this project is to develop novel statistical methods with clinical import to better understand genomic data and to assess important genomic and longitudinal biomarkers for the analysis and planning of screening regimens, clinical trials, and to better understand the etiology of the disease.